About me

What was your career and/or academic journey before EF?

I studied Physics as my Undergrad and for Master’s I studied CS and psychology across 3 European Universities in France, Spain, and Norway. At the same time, I started my research work, completely individually, I tried to do school projects using innovative methods that later could become papers, and they did! At that time I also interned in a few labs, namely in KU Leuven (Belgium) and Uni of Oxford (UK). These were experimental psychology labs, but what I did there is helping them to analyze experimental data using Data Science and Machine Learning techniques. In the end, I also worked on industrial projects, and after all, at my graduation, I had 14 papers published, quite a good number as for a pre-PhD person!

That, and excess of ambition helped me to make my way to the Facebook AI Research lab right after graduation. It was in their headquarter in Menlo Park, in one building with Mark Zuckerberg! Amazing opportunity and amazing people. I continued my research path, working in Vision & Language team where we concentrated on Multimodal AI – a novel type of algorithms that can "see" and "read" at the same time, and respond in natural language. Typical use of it, is, for example, by showing an image, ask AI "what sign is it?" or "what time does the clock show?". Intuitively, you could guess that it sounds like something visually impaired people would ask, and you will be right! It was used, first of all, for accessibility features to help people understand the world.

Why did you decide to join EF and become a founder?

During my academic time, I was surprised to observe how big the discrepancy between application and research is. Most of the algos we use today are stuff that being researched 5+ years ago! And you can imagine, for someone who has visibility into 5+ years in the future, how painful it is to use outdated tech and struggle around problems in daily life which I just read how to solve in a paper. Hence, my main motivation was to start applying the knowledge and skill I got to real-life problems (which I indeed do now at Suggestr).

There was also a personal aspect to it, as I realized that I can see where I will be after 10-15 years in academic research, and I wasn't too excited about it. That's the only life I have, and I know I'm able for more than passing a peer-review after 2 revisions :) An entrepreneurial career is much more scalable and fundamental than any "craft" I can possibly choose.

Did you have to give up anything to join the program?

After research at FB I could go to a top US school to do PhD, Also, I had offers from Google and Apple, but instead, even before joining a program, I left everything to do a stupid thing – try to hustle and build something from scratch. In that way, I had a couple of my first products built, including a deepfake app, a platform for practicing English, an interactive AI course, and a community of AI experts which now grew to 20,000+ members. But I wasn't happy with it, because while working on my own I was very limited. I gave a shot to EF to increase the scale of the game I play, and it was the best decision I ever made!

Why did you decide your co-founder was the right person to build your business with?

Adi is one of the smartest people I know. And when I say smart, I mean not "knows a lot of facts of formulas", I mean he has a remarkable thinking process, problem-solving ability, and communication. No skills matter to me more than that. People like him can change the world.

Are you exploring any unusual/novel technologies with your company?

No, why? We just do "git copy; python run".
But actually, since I know something, we do utilize it, and as a result, the recommendation engine we built is on par with top systems at Amazon, Facebook, TikTok (we even call it sometimes "Facebook-quality product recommendations", haha, just don't tell Facebook) and is nothing like what is available in the market to small and medium online businesses. We draw this line because of the exact same discrepancy I mentioned earlier, while huge companies can afford to be innovative and have the whole research labs which help them with that, the small guys are left to enjoy outdated basic algorithms created for mass-market. Most of them use the same approach called Collaborative Filtering developed in 1992 (we don't even have a single person in a company older than that). We say: screw collaborative filtering, let's build up from our belief how the perfect recommendation algorithm should work. It should understand what it recommends, right? Otherwise how it can do meaningful recommendations by just SKU number. So we applied Multimodal AI and transformers, which are frontier of research, to, first of all, enable the algorithm to understand the world by looking at the product images and descriptions, and only after that start doing recommendations. Surprisingly, just this additional step was enough to beat 95% of the competition. And the rest 5% is why I go to work today, and not chilling on the beach.

What are some of your notable achievements and failures to date?

Well, as for today, we made it to the homepage of the App Store, which is quite an achievement in our domain. Also, even though we already generated around $40K for our clients, just recently we received our own first revenue, which is a huge milestone for a company.

The biggest failure was when we broke a store of our very first client with our scrappy MVP. The biggest learning – don't break customer's stores.

Do you have any other thoughts on entrepreneurship and/or your EF experience you'd like to share?

I definitely enjoy the journey more than any imaginable outcome. Even though, you know, how people say "I left my 9 to 5 job, so now I work 24/7" is painfully true, it's still worth it. As for me, what makes it enjoyable is the feeling of ownership of my actions and their outcomes, and, of course, the people I do it with. I enjoy discussing business (and even arguing) with Adi, and I will do my best to preserve this intellectual culture with new team members and new investors.